On "starve the beast"

While browsing Greg's blog (see previous post) I also noted his item on "starve the beast"--the idea, popular with conservatives, that cutting taxes forces public spending lower. He says that a recent column by Paul Krugman implicitly endorses the theory, much as Paul may deplore its application, since it argues that the Bush tax cuts will tie the hands of the next administration. (Yes! That is what they were supposed to do, the White House would say.)

Perhaps I am misunderstanding, but Greg's point seems to be that Paul has somehow contradicted himself by endorsing "starve the beast". How so? One can perfectly well believe that the "starve the beast"/"poison pill" theory is true, and still deplore big tax cuts, so long as you also think that big increases in public spending are warranted (as Paul does). Indeed, the theory, if true, is the best reason for liberals to oppose tax cuts, just as for conservatives it is the best reason (or maybe second-best) to advocate them.

Question: Is the "starve the beast"/"poison pill" theory in fact correct? There is evidence pointing both ways--including the fact that Obama is promising to extend almost all of the Bush tax cuts, which would tend to support the idea. Still, on the whole I don't buy it. Here is a piece [pdf] on the subject I did for National Journal last year, which says why. It references a study by Christina and David Romer, which although certainly not the last word on the subject is for now, I think, the most authoritative word. The abstract of the Romers' study says this:

The hypothesis that decreases in taxes reduce future government spending is often cited as a reason for cutting taxes. However, because taxes change for many reasons, examinations of the relationship between overall measures of taxation and subsequent spending are plagued by problems of reverse causation and omitted variable bias. To deal with these problems, this paper examines the behavior of government expenditures following legislated tax changes that narrative sources suggest are largely uncorrelated with other factors affecting spending. The results provide no support for the hypothesis that tax cuts restrain government spending; indeed, they suggest that tax cuts may actually increase spending. The results also indicate that the main effect of tax cuts on the government budget is to induce subsequent legislated tax increases. Examination of four episodes of major tax cuts reinforces these conclusions.

I hope the Romers are right, as my NJ column explains:

The worst thing about "starve the beast" is the idea that a straightforward argument for low taxes and spending cannot be pressed successfully. You have to cut taxes, which the voters will like, and let them think they can have high spending, too. Later, if all goes according to plan, they will see they were mistaken. It is a strategy based not only on outwitting the Democrats, but on outwitting the electorate as well. It would be a pity if it worked.